Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/nihui/ncnn-webassembly-yolov5
Deploy YOLOv5 in your web browser with ncnn and webassembly
https://github.com/nihui/ncnn-webassembly-yolov5
artificial-intelligence chrome cross-platform edge firefox ncnn neural-network object-detection webassembly yolov5
Last synced: 12 days ago
JSON representation
Deploy YOLOv5 in your web browser with ncnn and webassembly
- Host: GitHub
- URL: https://github.com/nihui/ncnn-webassembly-yolov5
- Owner: nihui
- Created: 2020-11-19T09:35:51.000Z (almost 4 years ago)
- Default Branch: master
- Last Pushed: 2023-02-27T03:42:58.000Z (over 1 year ago)
- Last Synced: 2024-10-23T05:32:54.004Z (21 days ago)
- Topics: artificial-intelligence, chrome, cross-platform, edge, firefox, ncnn, neural-network, object-detection, webassembly, yolov5
- Language: C++
- Homepage: https://nihui.github.io/ncnn-webassembly-yolov5/
- Size: 8.06 MB
- Stars: 140
- Watchers: 4
- Forks: 30
- Open Issues: 10
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
- awesome-ncnn - ncnn-webassembly-yolov5
- awesome-yolo-object-detection - nihui/ncnn-webassembly-yolov5 - webassembly-yolov5?style=social"/> : Deploy YOLOv5 in your web browser with ncnn and webassembly. (Other Versions of YOLO)
- awesome-yolo-object-detection - nihui/ncnn-webassembly-yolov5 - webassembly-yolov5?style=social"/> : Deploy YOLOv5 in your web browser with ncnn and webassembly. (Other Versions of YOLO)
README
# ncnn-webassembly-yolov5
open https://nihui.github.io/ncnn-webassembly-yolov5 and enjoy
# build and deploy
Prior to starting, make sure you have `cmake` installed.
1. Clone the project - all other steps are to be performed inside the project directory.
```
git clone https://github.com/nihui/ncnn-webassembly-yolov5.git
cd ncnn-webassembly-yolov5
```
2. Install emscripten
```shell
git clone https://github.com/emscripten-core/emsdk.git
cd emsdk
./emsdk install 3.1.28
./emsdk activate 3.1.28
source emsdk_env.sh
```
3. Download and extract ncnn webassembly package
```shell
cd .. #Return to the root folder
wget https://github.com/Tencent/ncnn/releases/download/20230223/ncnn-20230223-webassembly.zip
unzip ncnn-20230223-webassembly.zip
```
4. Build four WASM feature variants
```shell
# The build folder is part of the bundle extracted in the previous step
cd build
cmake -DCMAKE_TOOLCHAIN_FILE=$EMSDK/upstream/emscripten/cmake/Modules/Platform/Emscripten.cmake -DWASM_FEATURE=basic ..
make -j4
cmake -DCMAKE_TOOLCHAIN_FILE=$EMSDK/upstream/emscripten/cmake/Modules/Platform/Emscripten.cmake -DWASM_FEATURE=simd ..
make -j4
cmake -DCMAKE_TOOLCHAIN_FILE=$EMSDK/upstream/emscripten/cmake/Modules/Platform/Emscripten.cmake -DWASM_FEATURE=threads ..
make -j4
cmake -DCMAKE_TOOLCHAIN_FILE=$EMSDK/upstream/emscripten/cmake/Modules/Platform/Emscripten.cmake -DWASM_FEATURE=simd-threads ..
make -j4
```
5. Deploy the *.data *.js *.wasm and *.html files to your web server
If you want to run this locally, create a folder named `deploy` in the root project directory and copy the following assets into it:
```
# deploy files
deploy/
├── index.html
├── yolov5-basic.data
├── yolov5-basic.js
├── yolov5-basic.wasm
├── yolov5-simd.data
├── yolov5-simd.js
├── yolov5-simd-threads.data
├── yolov5-simd-threads.js
├── yolov5-simd-threads.wasm
├── yolov5-simd-threads.worker.js
├── yolov5-simd.wasm
├── yolov5-threads.data
├── yolov5-threads.js
├── yolov5-threads.wasm
├── yolov5-threads.worker.js
└── wasmFeatureDetect.js
```
6. Deploy local server(python3 as a example)
```
python3 -m http.server --directory deploy
```
7. Access local server(chrome as a example)
```
# launch chrome browser, enter following command to address bar and press ENTER:
chrome://flags/#unsafely-treat-insecure-origin-as-secure
# enter following keyword to "Search flags" and press ENTER:
"insecure origins"
you will find "Insecure origins treated as secure" key
#enter local server url and click right side dropdown list, select "Enabled"
url example: http://192.168.1.100:8000
#relaunch chrome browser and access http://192.168.1.100:8000 (replace 192.168.1.100 with your local ip)
```